Line-related bacteremias associated with long-term catheters (Hickman/Broviac, Groshong, pulmonary artery catheter, and peripherally inserted central catheter) are being encountered regularly now that these catheters are used so commonly. Earlier studies indicated that these catheter-related episodes did not necessitate removing the offending line if the infecting organism was a coagulase-negative Staphylococcus, most commonly S. epidermidis. Antibiotic treatment with the line left in place generally eradi-cated the organism. For other organisms, however, particularly gram-negative rods, this issue has not yet been resolved. We would like to examine whether removal of the catheter results in more reliable initial clearance of the gram- negative rod than nonremoval, and also whether the incidence of recurrent infection with the same organism increases in patients in whom the catheter was left in place. Approximately 70 episodes of catheter-related gram-negative bacteremia seen over the past 2.5 years are being examined to determine whether there are any significant differences between patients with catheters removed and those with catheters left in place. Patient data on these patients is currently being reviewed. In addition, more patients will need to be accrued in order to have statistically evaluative findings. This work is still in progress, and should be ready for analysis this year.